Now that they've fixed the mail server and I can post again...

Yes, its sad to see talented modellers go but to be honest, for those that have simply got bored with modelling for DoD I can sympathise.

I'm lucky in that I don't have a life, hence I dont have to feel that the real world has a better reason to drag me away from my computer.

However I must confess that since 3.1 DoD has turned me off a little from modelling a great deal.

I think many people look at the newer games coming out with shiney graphics and great frame rates, then look at DoD and think - what the HECK am I doing with this old piece of ****?

For me, modelling for DoD is a masochistic high. I love nothing better than slaving away for 3 months on a set of player models - hand tweaking every vertex and bone aligment in the pursuit of smoother curved joints.

Not everyone shares my perversion but making a technology do something its not meant to do is all part of the fun.

Even so, with the work I've done on HLMV I've discovered a lot about the internals of Half-Life an realised on a technical level that most of the Half-Life engine as it stands now is a hack.

Bear in mind, its not the DoD team who hacked the engine, Valve did and its things like adding transparency and additive modes which have had some detrimental effects. I wont go over it technically but trust me, to get HLMV to do it is messy so I fear what the internals of Half-Life must look like.

Its also fun to watch the trend since 3.1 came out. MSA was full of people pushing poly-counts and levels of detail higher and higher. The Half-Life engine was pretty fast and could handle it at that time.

However with all the superfluous junk that Valve added to 1.0 they have added a lot of internal "choke" to the HL-engine which means that a lot of CPU time is wasted which makes out high poly stuff a bad thing.

Its therefore interesting to note that since 1.0 came out we went through a stage of cursing the submodel problem, then silence as we discovered our high-poly stuff ran REALLY badly in 1.0 to the situation now where we're release LOW POLY stuff like a 20 triangle v_model garand just to try and eek back a bit of performance.

It is sad to see MSA in decline but its encouraging to see new members who want to get involved asking for help at the same time.

For an engine thats been around since 1999, HL 1 has done well but I think Valve have with the DLL updates to 1.0 put the pillow over the face of their own baby. The DoD team is not to blame, Valve have just tried to ring the last drops out of their technology and kind of killed it by doing too much.

But for those of you leaving you've contributed and learnt a lot while being here so don't let it go to waste. If you need a change, find other mods, other game engines and stay involved because there are too many games out there that are bland and repetitive.

We need modders and modelers who are prepared to rise to a challenge, to push engines as far as they can go and to add some creativity back into computer games.

- Jed

word

Very much agreed on the sentiment about continuing. I don't know enough about the technicalities to agree on the engine stuff, though.

Well said.

its pretty obvious that this community is slowly coming to an end... course, i knew jack **** when i entered this community. and now i know the ins and outs of it all.

theres always more to learn, and hopefully well see everyone here at another forum.. most likely the COD forums. and well pick it up again, and learn from scratch until weve reached that level of skill again.

as opposed to before when a pack could take me weeks, can now be completed in a sinlge week, with the knowledge i know now. its been fun, but all good things come to an end eventually. :\

Well said.

I've been playing since 1.3b, but I entered the modding scene with my first reskin just a few days before 1.0 came out, but since then I've learned an awful lot about skinning and QC files and I have you guys to thank! :)

I hope with 1.1 some things will be fixed and performance increased, so this community can get back to doing what it does best, offering the highest quality addons for any mod out there. :D

if DoD was as widespread as CS we would have alot more modellers/skinners, its just a matter of WHO is interested and has the patience to deal with all opinions in MSA. but i must say, right now, we've got the toughest, cool-headed people around to keep this forum going.

I model. ;o

http://www.boomspeed.com/tbfrog/xm177-6.jpg

I think 1.1 may get things rolling again, reducing bugs and fps laggers that should have been out of 1.0 in the first place. Not to mention no more submodels so someone without a Computer Science Degree from MIT can use parts of a model set without using all or none, and best of all for me, our old friend the 2/3.x crosshair.

Originally posted by speedealer
word
up

Very well said Jed. I too like others have arrived to the scene late, but it doesn't mean I won't stop from doing what I love. I think over time as others realize that customization is the key in keeping an old but fun engine enjoyable, they too will join the community. From experience, it's obvious that being able to play with models/animations/coding/sprites that you created just makes the experience worth while. Personally I myself know I will be around this community for a long time untill I master Half-Life's player animation attributes. Even then when I do master it I will continue supporting this most memorable mod even way into HL2, but never forget that there will always be people who won't be able to enjoy the fine luxuries of a high end pc specified game. This engine has come a long way and will continue to get better with the many new discovered talent of the people who still enjoy it.

Guess who might make a comeback soon...:)

Real soon....

Originally posted by Cs42886
Guess who might make a comeback soon...:)

Real soon....

Nomad and Slinky?

LOL

:EDIT: Oh and mike, Hallock is Nomad.. :o

Originally posted by Mike[1stid]
Nomad and Slinky?


ROFL

Eh heh, my 20 poly garand wasn't to squeeze out performance, I'm running quite fine with my 2.1 ghz Athlon 2600+ and Radeon 9800, thank you hahahah.

I want to see more ORIGINAL content, instead of hackjobs consisting of a 100000 poly colt hanging on the holster. Look, there's a huge difference between "High-poly for detail" and "High poly for the sake of having lots and lots of triangles", and a lot of hackjobbers don't seem to realize that. A crappy modeller can make a 30,000 or whatever poly MG42 look good. A great modeller can make a <1000 poly MG42 and make it look good. Both ways it looks good, but with one your computer's going to run at the speed of a fat man on crack. Get it?

High poly != Good Model
Low Poly != Crappy Model
Good Model = Good use of polys and nice texturing/etc.

Pushing the envelope is a good thing, but waiting for an engine to change just so you can waste polys is a waste of time. Learn how to use polys effectively and you'll be much better off.

word to that too. i hate all those awesome looking model packs with hacked pistols and canteens and whatever that have huge and numerous textures (that aren't sized correctly EVER) and they slow down my computer like a *******, especially if those textures aren't sized to be divisible by eight which always seems to be the case for a good lot of custom work. i'm not bashing anyone.. well yes i am, but only the fact that people overlook that important texture detail when recompiling. and i understand those textures are resized when the new model is compiled but they had to be in eights at some point, going the extra mile to make them all eights again would probably kick my (and everyone else's) fps up by ten or twenty frames.

$cliptotextures

wow, you are awesome. thanks. and i even know what to do with that.

Your welcome..:) and thanks...:)

It's true though, i have seen many a good model been released on these forums where the $cliptotextures command has not been used when the model has been compiled, so the textures end up being odd sizes. It does not happen that often now though, because the new decompiler adds the command to the qc file, when you decompile a model. But i always use the $cliptotextures command, because sometime ago through testing i found that it can make your textures blurred if you don't, and this combined with the halflife engine scaling the cropped texture ingame aswell, only added to making it look even worse IMO.

So i say it again.

$cliptotextures

And as this thread swings wildly off on a tangent...

About $cliptotextures and power-of-2 sizes...

$cliptotextures only prevents the compiler from cropping the images down to what portion is contained within the boundries of the UV Map.

When DoD/HL loads your model all textures are checked to see if their width and height are power-of-2 and if not, are scaled UP to the next nearest dimension that is.

So, say your width is 63, it gets scaled to 64. If its 200, its scaled to 256 and in the case of 257, scaled to 512.

As you can see, for larger numbers the scaling is greater and hence blurring of your textures gets worse.

The scaling is a requirement of your graphics card 3D hardware - its only newer cards which can handlle non-power-of-2 sizes.

The point being is, using $cliptotextures or not doesnt matter because if the final dimensions are not power-of-2 it will be scaled and in the case of some textures will result in MORE graphic card memory being used than needed. (257x257 being scaled to 512x512).

Solution?

If your textures dimensions are just under a power-of-2 in any direction PAD out the texture up to it using space.

If your just over a power-of-2 in any dimension, scaled DOWN your texture to the nearest value and re-align your UV Map.

Your textures will look better and conserve graphic memory.

- Jed

Originally posted by Fiblah
LOL

:EDIT: Oh and mike, Hallock is Nomad.. :o

I know. :)